Photoelectric catalytic degradation of methylene blue by C60-modified TiO2 nanotube array

Applied Catalysis B: Environmental 89 (2009) 425–431

Contents lists available at ScienceDirect

Applied Catalysis B: Environmental journal homepage: www.elsevier.com/locate/apcatb

Photoelectric catalytic degradation of methylene blue by C60-modified TiO2 nanotube array Jie Lin, Ruilong Zong, Mi Zhou, Yongfa Zhu * Department of Chemistry, Tsinghua University, Beijing, 100084, PR China

A R T I C L E I N F O

A B S T R A C T

Article history: Received 2 September 2008 Received in revised form 20 December 2008 Accepted 30 December 2008 Available online 15 January 2009

Fullerene (C60)-modified TiO2 nanotube array (TNA) was prepared by the electrophoresis deposition technique. The as-prepared samples showed the high efficiency for the photoelectric catalytic (PEC) degradation of nonbiodegradable azodyes methylene blue (MB). The highest PEC activity of C60-modified TNA (TNA/C60) was achieved at a lower bias potential (4.0 V), which was 2.3 times of the highest activity of TNA at 5.0 V. The high PEC activity came from the synergetic effect between C60 and TiO2, which promoted the charge separation, influenced the charge distribution of the electrical double layer and reduced the impedances of the Helemholtz and depletion layers. Moreover, the oxidation of MB was a quick process during the PEC degradation, and the process began with the oxidation of the dimethylamino group, which was different from the photocatalytic (PC) process began with the oxidation of S atom; MB was mineralized completely during PEC degradation. ß 2009 Published by Elsevier B.V.

Keywords: Photoelectrical catalysis Fullerene TiO2 nanotube array Methylene blue

1. Introduction Photocatalytic processes are a promising class of advanced oxidation technologies used for environmental remediation [1,2]. Among the photocatalysts, TiO2 has been intensively investigated for the complete degradation of recalcitrant organic pollutants [2– 4], because it is easily available, nontoxic, low-cost and chemically stable. However, TiO2 has two typical shortcomings to render its wide application in practice: the difficult separation of TiO2 from aqueous phase and relatively low quantum yield due to the rapid recombination of charge carriers. Now, the separation of photocatalyst from the solution suspension can be fulfilled by the immobilization of TiO2 particles. Immobilization of TiO2 on the supporter, especially on the conducting substrate, not only can eliminate the need for separation of photocatalyst from the solution suspension, but also can provide more advantages, such as using electrochemical technique to study and cooperate with photocatalytic process expediently [5,6]. For the advanced oxidation application, TNA has been considered as the most suitable way to achieve larger enhancement of surface area without an increase in the geometric area [7]. Among the fabrication methods of TNA, the electrochemical synthesis method shows its advantages of good mechanical adhesion strength and

* Corresponding author. Tel.: +86 10 62783586; fax: +86 10 62787601. E-mail address: [email protected] (Y. Zhu). 0926-3373/$ – see front matter ß 2009 Published by Elsevier B.V. doi:10.1016/j.apcatb.2008.12.025

electronic conductivity since it directly grows from the titanium metal substrate [8]. In addition, the thickness and morphology of such TiO2 film are easily controlled. The high degree of recombination between photogenerated electrons and holes in semiconductor particles is a major limiting factor for photodegradation process [9]. Among the ways to enhance the separation of photogenerated charge carriers, the application of a low bias is an effective method. The bias drives the photogenerated electron to counter electrode in the PEC process, which could counteract the charge recombination process [10–12]. C60 have attracted extensive attentions for their various interesting properties due to their delocalized conjugated structures and electron-accepting ability. One of the most remarkable properties of C60 in electron-transfer processes is that it can efficiently arouse a rapid photoinduced charge separation and a relatively slow charge recombination [13]. Thus, the combination of photocatalysts and C60 may provide an ideal system to achieve an enhanced charge separation by photoinduced electron transfer. Some of the fullerene-donor linked molecules on an electrode exhibited excellent photovoltaic effects upon photo-irradiation [14–16]. Our previous study showed that the photocatalytic activities of Bi2WO6 were improved remarkably after C60 modification [17], the combination of C60 and TNA may be an ideal system for the PEC degradation of organic pollutant. In this work, the enhanced PEC activity of TNA/C60 was reported. A series of experiments aimed to investigate the relationship between the C60 modification and the PEC activity

426

J. Lin et al. / Applied Catalysis B: Environmental 89 (2009) 425–431

were described. It showed that the enhanced activity was due to the synergistic effect of C60 and TNA. Moreover, the intermediates detected during the PEC process showed a different degradation path of MB from the PC process. 2. Experimental 2.1. Materials and preparation All chemicals were analytical grade reagents and used without further treatment. Electrolyte was freshly prepared from deionized water. After chemical polishing, titanium foil (thickness about 250 mm, purity 99.4%, Beijing Cuibolin Non-Ferrous Technology Developing Co., Ltd.) was subjected to potentiostatic anodization in an electrochemical anodization cell with a platinum cathode in a 0.5 wt% HF + 1 M H3PO4 electrolyte at ambient temperature. The potential of 20 V was applied for 30 min. Then the samples were rinsed with deionized water and annealed at 450 8C for 24 h. The TNA/C60 samples were prepared using electrodeposition method as in Ref. [18]. The electrolytic suspension was prepared by adding C60 (purity 99.9%, Peking University, PR China) to a mixture of acetonitrile and toluene (3:1, v/v). The electrodeposition was carried out potentiostatically using a CHI660B electrochemical system (Shanghai, China), with the TNA film as working electrode, a platinum wire as counter electrode and a standard calomel electrode (SCE) as reference electrode, respectively. The coverage of C60 on the TNA film was estimated by the charge passed (after the correction for solvent blank) during the electrodeposition. 2.2. Characterization The structures of the TNA and TNA/C60 samples were characterized by XRD (Rigaku D/MAX-2500 X-ray powder diffractometer), using graphite monochromatized Cu Ka radiation (l = 0.154 nm). A 0.02 step in 2u/count, beam voltage of 40 kV and beam current of 300 mA were used. The phase composition of the samples was determined by Microscopic Confocal Raman Spectrometer (Renishaw, RM2000) using 632.8 nm as the exciting light source. Spectra were collected in the range of 1000–200 cm1 with a resolution of 1 cm1. The morphologies and microstructures of the as-prepared samples were observed using a field emission scanning electron microscope (FE-SEM, LEO-1530) at 10 kV and a Tecnai TF20 high-resolution transmission electron microscope (HRTEM) operated at an accelerating voltage of 200 kV. Chemical characterization of the sample surface was recorded with scanning X-ray photoelectron spectroscopy (XPS, ULVAC-PHI, Quantera). The beam voltage was 3.0 kV, and the energy of Ar ion beam was 1.0 keV. The sputtering rate was approximately 9.5 nm/min for a thermally oxidized SiO2 thin film. The binding energies were normalized to the signal for adventitious carbon at 284.8 eV.

were further fitted and interpreted by Zsimpwin software. The catalytic activities of the samples were all evaluated by the removal of MB dye (with an initial concentration of 10 mg/L). The changes of MB concentration were monitored by the variations in absorption intensity at 660 nm using a UV–vis spectrometer (Hitachi U-3010). The mineralization of the dye was followed by measuring the total organic carbon (TOC) concentration, utilizing a Shimadzu Corporation TOC-V wp Analyzer. 2.4. Analyses of MB intermediates To analyze the PEC process of MB, the original concentration of MB was increased to 50 mg/L, and the neutral products were enriched by 140 times. Before the analysis, the samples were filtered through millipore discs of 0.45 mm to protect the chromatographic column. HPLC monitoring was carried out using a UV absorbance detector (K 2501) operated at 280 nm coupled to a Venusil XBP-C18 (Agela Technologies Inc.) column. According to the literature [19], the reversed-phase eluent of pH 3 buffer and methanol (45:55, v/v) were used for aqueous solution, and water and methanol (40:60, v/v) was used for enriched neutral products. The neutral intermediates were finally identified by LC/MS (Thermo Fisher, LTQ). 3. Results and discussion 3.1. Photoelectric properties The photoelectric properties of the TNA and TNA/C60 samples were evaluated by the photocurrent at bias 0 V and 1 V (Fig. 1). The amount of C60 on the TiO2 film was expressed by the charge passing during the electrodeposition. It can be seen that the photocurrent was enhanced with the deposition charge and reached the maximum at 0.123 mC. The further increase decreased the photocurrent and finally reached its balance even the deposition charge increased to 0.41 mC. The highest photocurrent of TNA/C60 was 30% higher at 0 V and 40% higher at 1 V than that of TNA, respectively. To investigate the influence of C60 modification on the photoelectric property, the EIS technology was used to study the solid/electrolyte interfaces of the TNA and TNA/C60 samples. According to conventional double-layer theories, the electrical double layer at the solid electrode behaved as a frequency distribution impedance instead of a pure capacitance due to the surface heterogeneity. When the charge transfer reaction occurred,

2.3. Photoelectric properties and photoelectrocatalytic activities All electrochemical and photoelectric studies were performed on a CHI660B electrochemical system (Shanghai, China) using a standard three-electrode cell with a working electrode (20 mm  45 mm), a platinum wire counter electrode, and a SCE reference electrode. Photoelectrochemical properties were measured with an 18 W germicidal lamp (l = 254 nm, Institute of Electric Light Source, Beijing). Unless otherwise stated, the intensity of light at the film electrode was 1.64 mW/cm2 at the wavelength of 254 nm, and 0.1 M Na2SO4 electrolyte was used. The photocurrents were measured in the potential range of 0.3 to 1.0 V. The electrochemical impedance spectroscopies (EIS) were carried out at the open circuit potential. A sinusoidal ac perturbation of 5 mV was applied to the electrode over the frequency range of 0.05–105 Hz. The EIS spectra

Fig. 1. Influence of deposition charge of C60 on the photoresponse for the TNA/C60 film.

J. Lin et al. / Applied Catalysis B: Environmental 89 (2009) 425–431

the Nyquist plot was a semicircle; when semi-infinite diffusion was the rate-determining step, a linear with a slope of 458 appeared [20]. In our cases (Fig. 2a and b), only one semicircle on the EIS plane suggested charge transfer occurring, and the equivalent circuits were shown in Fig. 2c and d [21]: R1, solution resistance; R2, electric charge transfer resistance, corresponding to the Helmholtz layer; R3, corresponding to the depleting layer; Q, the constant phase elements (CPE) of the inner layer of the TNA/C60 sample. Using the equivalent circuit, the impedance fitting values were shown in Fig. 2e and f. The electric charge transfer resistance (R2) and the depleting layer resistance (R3) under dark field decreased significantly with C60 modification and reached the balance when deposition charge of C60 increased to 0.082 mC.

427

The C60 on the surface increased the process of charge separation, influenced the distribution of the electrical double layer, and promoted the electron transfer, so that the electric charge transfer resistance (R2) decreased. The charge distribution of electrical double layer also impacted the depleting layer, so that R3 also decreased. With the increasing coverage of C60 on the surface of TNA film, the process of charge transfer reached the balance, and R2 and R3 were also invariable. Under UV irradiation, the formation of photoinduced electron– hole pair reduced the resistance of depleting layer (R3) by one order. Different from dark field, the Helmholtz layer was mainly occupied by the photoinduced electron–hole pair, so R2 was also reduced by on order. The electron-transfer characterization of C60

Fig. 2. Impedance spectra of TNA and TNA/C60 films (a) under dark and (b) under UV irradiation. Equivalent circuit supposed of TNA and TNA/C60 films (c) under dark and (d) under UV irradiation. Impedance fitting values of the TNA/C60 samples with different disposition charges (e) under dark and (f) under UV irradiation.

428

J. Lin et al. / Applied Catalysis B: Environmental 89 (2009) 425–431

increased the charge separation process of the Helmholtz layer and changed the charge distribution of depleting layer availably. The smaller arc radius on EIS Nynquist plot of TNA/C60 film under UV irradiation meant an effective separation of photogenerated electron–hole pairs and fast interfacial charge transfer occurred [17]. On the other hand, the C60 on the TNA surface reduced the UV absorption and the contacting area between the TiO2 and solution, which decreased the utilization efficiency of light. The enhanced charge separation process and reduced efficiency of light competed in our system, which caused the photoelectric properties increased initially then decreased to the balance, and the highest photoresponse was obtained at the deposition charge of 0.123 mC. 3.2. Photoelectrocatalytic activities The PEC activities of TNA and TNA/C60 (deposition charge of 0.123 mC) were evaluated by the degradation of MB under UV irradiation. Only 1% of MB was adsorbed on the TNA/C60 film after 2 h, so that the adsorption was not the main factor on the PEC process. According to the pseudo-first-order kinetic, the kinetic constants with the different biases were shown in Fig. 3a. As the electric catalytic (EC) degradation did not obey the pseudo-firstorder kinetic, the percent of color removal after 90 min reaction was also given in Fig. 3b. It was seen that 12% of MB was removed at 6.0 V by EC process and 18.4% of MB was removed only under UV

irradiation, so that the EC process and UV irradiation were not effective routes to degrade MB. During the PEC process, the reaction rate increased initially then decreased with the increase of bias, which was similar to our previous study [22]. The low electrical bias between the anode and cathode, driving the photogenerated electron to counter electrode, could counteract the charge recombination process and improve the activity of TiO2 [10,11]. The curve of cyclic voltammetry showed direct electric oxidation of MB on the TNA/C60 sample occurred above 3.2 V, and the oxygen evolution reaction occurred above 3.7 V (Support Information Figure S1). With the bias in the range of 2–3 V, the influence of bias on the separation of photoinduced electron–hole pair reached the balance, and the PEC activity could be attributed to the absolute amount of photoinduced electron–hole pair, which was constant under the UV irradiation of the same intensity. When the bias reached 3.2 V, the photoinduced-hole oxidation and the direct electro-oxidation of MB occurred together, and the photodegradation and electrochemical polymerization of MB competed on the sample surface. When the potential exceeded 3.7 V, the more active species such as hydroxyl radicals, H2O2, or O3 could be produced [22], which lead to the indirect oxidation of MB. The mutual control of the three sides determined the maximum degradation rate reached at 4.0 V. With the bias further increasing, the process of electrochemical polymerization exceeded that of hole-oxidation and hydroxyl radicals’ indirect oxidation, the polymer accumulated on the surface prevented the current conduction and decreased the degradation rate. The presence of C60 reduced the resistance of Helmholtz layer and depleting layer, so the direct electro-oxidation and indirect oxidation of MB occurred at a lower bias. The presence of C60 also improved the separation of photoinduced electron–hole pairs and electron transfer, causing the increase of the photoelectrical activity. In a word, the modification of C60 on the surface enhanced the PEC activity of TNA and the highest activity can be achieved at a lower bias (4.0 V). As all the experiments were repeated, TNA/C60 was stable during the PEC process. 3.3. Intermediates during the PEC degradation of MB

Fig. 3. The catalytic activity of TNA and TNA/C60 samples. (a) The pseudo-first-order kinetic constants of the PEC degradation of MB with TNA and TNA/C60 samples; (b) the color removal of MB within 1.5 h under PEC, EC, UV reaction, respectively.

To investigate the PEC process of MB, the composition of MB intermediates were detected by HPLC (Fig. 4). There were only two peaks separated from the chromatogram spectra of the original reaction solution. The previous study [19] indicated that two forms of MB existed in the solution (Support Information Scheme S2), so the two peaks at 4.5 and 4.6 min assigned to compounds II and I, respectively. It was clear that the compound II was degraded more quickly than the compound I during the PEC process, which suggested the compound II was more easily oxidized. To examine the process in detail, the neutral intermediates enriched by 140 times were also detected by HPLC (Fig. 4b) and further identified by LC/MS. The suggested structures of the intermediates based on the LC/MS results were shown in Table 1. The neutral intermediates separated at 11.2, 11.6 and 17.7 min accumulated during the PEC process (Fig. 4b). The fact that all the three intermediates had a cyclohexa-2, 5-dienone structure suggested that the PEC process started with the oxidation of the dimethylamino group. Thus the compound II was considered as the initial reactant oxidized on the TiO2 surface, so that the concentration of the compound II decreased more quickly. The minor intermediates separated at 8.0 and 8.5 min showed the primal oxidation at S atom, which was also observed at the initial step of MB degradation during the PC process [23]. These two intermediates suggested that a part of MB was degraded via the PC route. For the concentrations of all the neutral intermediates were low and the small molecules at 3.7 min did not accumulated markedly, it made evident that the PEC degradation was a quick

J. Lin et al. / Applied Catalysis B: Environmental 89 (2009) 425–431

429

Table 1 Suggested structures for the intermediates based on LC–MS results. Retention time, tR (min)

Structural formula

Molecular weight

HO 3S

3.7

O

174

Cl N

235

SO3H

Cl

202

O2N NO 2 Cl

128

O

7.4

Cl N

O S

Cl O

340

N 8.0

Cl N

O S

Cl N

369

N

337

N 8.5

Cl O2N

O S N

11.2

S

N

Fig. 4. Chromatograms of MB degradation products with the initial concentration of 50 ppm in photoelectrical system. (a) Original solution in methanol–buffer (55:45, v/v) eluent; (b) the neutral intermediates enriched by 140 times in methanol–water (60:40, v/v) eluent.

O S

O2N

O

274

N 14.2

O S

O2N

294

NO2 14.6

3.4. Structure of TNA/C60 film The XRD results (Support Information Figure S3) demonstrated that TiO2 nanotube held the anatase phase before and after C60 modification. No diffraction peak of C60 was detected and C60 maybe existed in microcrystal or amorphous form. In addition, TiO2 nanotube was oriented along [0 0 1] direction. The morphologies of the TNA and TNA/C60 were obtained by FESEM (Fig. 5a and b). The TNA was estimated 400 nm long, and the internal diameter was in the range of 50–70 nm. The morphology of TNA/C60 remained the same as that of TNA, and no aggregation of C60 was observed on the surface of the TNA/C60 sample. Further study by

256

N 11.6

process. The mineralization of dye was also studied by measuring TOC removal. The samples corresponding to the HPLC results at 0 h and 12 h were used. After PEC process for 12 h, the TOC concentration decreased from 22.25 ppm to 4.37 ppm, namely 80% of TOC was removed by PEC oxidation. This result was consistent with the HPLC result. In a word, the photoelectrical oxidation of MB was a quick process and mainly began with the oxidation of the dimethylamino group; MB was mineralized by 80% during PEC degradation for 12 h.

O

Cl O 2N

O S

Cl O

376

O

306

NO 2 17.7

N

O S NO2

430

J. Lin et al. / Applied Catalysis B: Environmental 89 (2009) 425–431

Fig. 5. Morphology of TNA and TNA/C60 samples: (a) FESEM image of TNA at the top view, the inset was that at the cross-section view; (b) FESEM image of TNA/C60 sample at the top view; (c) HRTEM image of the close-end of TNA/C60 sample lift-off from the film; (d) HRTEM image of the open-end of TNA/C60 sample with C60 clusters signed; (e) HRTEM image of the TNA sample.

HRTEM (Fig. 5c) showed the barrier layer (close-end) of TNA/C60 was structurally uniform with a lattice spacing of 0.377 nm corresponded to the (1 0 1) plane of anatase TiO2 and the [1 0 0] direction was vertical to the growth direction. At the open-end of the TNA/C60 (Fig. 5d), some particles with amorphous form dispersed on the TiO2 tube wall, and the average particle diameter was about 2 nm, which suggested that the C60 on the TiO2 tube wall existed in the cluster form. To confirm the existing form of C60 on the TiO2 surface, the pristine C60, TNA/C60 and fresh TNA samples were further characterized by Laser Raman spectroscopy (Fig. 6). The peaks of the TNA and TNA/C60 samples at 144, 397, 517 and 633 cm1

belonged to the vibration mode of anatase phase, which confirmed the TiO2 nanotube array held the anatase phase during the C60 modification. The two peaks of the pristine C60 at 1468 cm1 and 1571 cm1 were observed, which was corresponded to the Ag (2) and Hg (8) mode of C60 single crystal [24]. With Gaussian fitting, the wide peak of TNA/C60 around 1600 cm1 was fitted to three peaks at 1561, 1580 and 1623 cm1, which connected with Hg (8) modes of C60, the first-order G band and the disorder-induced D0 band of quasi-graphite structure, respectively [25]. Compared with the pristine C60, the peaks of Ag (2) and Hg (8) modes in the TNA/ C60 sample downshifted to 1460 and 1561 cm1, and the line widths increased observably. The downshift and increased line

J. Lin et al. / Applied Catalysis B: Environmental 89 (2009) 425–431

431

electric response. The enhanced charge separation process induced the TNA/C60 sample reached the highest PEC activity at a lower bias. Furthermore, the photoelectrical oxidation of MB was a quick process and began with the oxidation of the dimethylamino group on the TiO2 surface; 80% of MB was mineralized during PEC degradation for 12 h. Acknowledgements This work was partly supported by Chinese National Science Foundation (20673065) and National Basic Research Program of China (2007CB613303).

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.apcatb.2008.12.025. Fig. 6. Raman spectra for (a) TNA, (b) pristine C60 and (c) TNA/C60 samples. The inset was the comparison spectra of (b) and (c) in the range of 1400–1680 cm1 after background correction.

widths of Ag pentagonal pinch mode indicated the electron transfer from TNA to C60 [26]. The appearance of G band and D0 band was supposed due to the partly ordered arrangement of C60 on the surface of TiO2. The ordered arrangement of C60 increased the charge transfer and fluorescence by conjugation effect, so that Raman spectrum of the TNA/C60 sample displayed observable fluorescence background. Based on the XPS results of TNA/C60 sample (Support Information Figure S4), the existence of the peak at 282.3 eV in C 1s spectrum and the peak at 457.6 eV in Ti 2p spectrum confirmed that C and Ti had a weak chemical interaction, which was consistent with the Raman results. Based on the results above, the coverage of the C60 on the TNA/C60 sample with the deposition charge of 0.123 mC was estimated in Supporting Information Figure S5. In the TNA/C60 sample of the best photoresponse (the deposition charge was 0.123 mC), nearly 88% of the TiO2 tube wall was covered the C60 clusters. 4. Conclusions In summary, TiO2 nanotube array modified with C60 was formed by the electrodeposited method, and the synergetic effect between C60 and TNA enhanced charge separation process remarkably. C60 modification on the TNA surface with 88% coverage optimized the competition of enhanced charge separation process and reduced efficiency of light, and the sample obtained the highest photo-

References [1] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995) 69– 96. [2] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C 1 (2000) 1–21. [3] K.I. Hadjiivanov, D.K. Klissurski, Chem. Soc. Rev. 25 (1996) 61–69. [4] A. Heller, Acc. Chem. Res. 28 (1995) 503–508. [5] D.L. Jiang, S.Q. Zhang, H.J. Zhao, Environ. Sci. Technol 41 (2007) 303–308. [6] W.X. Dai, X.X. Wang, P. Liu, Y.M. Xu, G.H. Li, X.Z. Fu, J. Phys. Chem. B 110 (2006) 13470–13476. [7] Z.H. Zhang, Y. Yuan, G.Y. Shi, Y.J. Fang, L.H. Liang, H.C. Ding, L.T. Jin, Environ. Sci. Technol. 41 (2007) 6259–6263. [8] Y. Xie, Electrochim. Acta 51 (2006) 3399–3406. [9] A.L. Linsebigler, G. Lu, J.T. Yates Jr., Chem. Rev. 95 (1995) 735–758. [10] K. Vinodgopal, S. Hotchandani, P.V. Kamat, J. Phys. Chem. 97 (1993) 9040–9044. [11] K. Vinodgopal, P.V. Kamat, Sol. Energy Mater. Sol. Cells 38 (1995) 401–410. [12] X. Quan, S. Yang, X. Ruan, H. Zhao, Environ. Sci. Technol. 39 (2005) 3770– 3775. [13] G. Yu, J. Gao, J.C. Hummelen, F. Wudl, A.J. Heeger, Science 270 (1995) 1789–1791. [14] H. Hotta, S. Kang, T. Umeyama, Y. Matano, K. Yoshida, S. Isoda, H. Imahori, J. Phys. Chem. B 109 (2005) 5700–5706. [15] D. Hirayama, T. Yamashiro, K. Takimiya, Y. Aso, T. Otsubo, H. Norieda, H. Imahori, Y. Sakata, Chem. Lett. (2000) 570–571. [16] D.F. Liu, S.H. Yang, S.-T. Lee, J. Phys. Chem. C 112 (2008) 7110–7118. [17] S.B. Zhu, T.G. Xu, H.B. Fu, J.C. Zhao, Y.F. Zhu, Environ. Sci. Technol. 41 (2007) 6234– 6239. [18] G.K.R. Senadeera, V.P.S. Perera, Chin. J. Phys. 43 (2005) 384–390. [19] J.D. Donaldson, S.M. Grimes, N.G. Yasri, B. Wheals, J. Parrick, W.E. Errington, J. Chem. Technol. Biotechnol. 77 (2002) 756–760. [20] J.Q. Li, L. Zheng, L.P. Li, Y.Z. Xian, L.T. Jin, J. Hazard. Mater. 139 (2007) 72–78. [21] W.P. Gomes, D. Vanmaekelbergh, Electrochim. Acta 41 (1996) 967–973. [22] X. Zhao, Y.F. Zhu, Environ. Sci. Technol. 41 (2006) 3367–3372. [23] A. Houas, H. Lachheb, M. Ksibi, E. Elalour, C. Guillard, J.M. Herrmann, Appl. Catal. B: Environ. 31 (2001) 145–157. [24] P. Bowmar, W. Hayes, M. Kurmoo, P.A. Pattenden, M.A. Green, P. Day, K. Kikuchi, J. Phys. Condens. Matter 6 (1994) 3161–3170. [25] L.G. Canc¸ado, A. Jorio, M.A. Pimenta, Phys. Rev. B 76 (2007) 064304. [26] Y. Zhang, Y. Du, J.R. Shapley, M.J. Weaver, Chem. Phys. Lett. 205 (1993) 508–514.